Integrating 3DGS novel view synthesis and CFD for modeling bionic robotic fish from multi view imagery

Simulating hydrodynamic characteristics of bio-inspired robotic fish through reverse engineering is challenging due to the difficulty in obtaining accurate 3D models efficiently for CFD analysis. This study proposes a novel framework integrating 3D Gaussian Splatting (3DGS) reconstruction from multi-view 2D images with CFD simulations. Our approach replaces costly 3D scanning by reconstructing high-fidelity models directly from multi-view camera data, enabling efficient hydrodynamic analysis using k−ϵ and k−ω SST turbulence models. Results demonstrate that 3DGS provides geometrically accurate CFD inputs while significantly reducing model preparation time. This framework offers a cost-effective solution for bio-inspired robot design and can be extended to other underwater systems for data-driven hydrodynamic evaluation.